Walberla: a Multi-Physics Open-Source Software Framework for Scalable and Efficient CFD

Kemmler S, Köstler H (2023)

Publication Language: English

Publication Type: Conference contribution, Abstract of lecture

Publication year: 2023

Event location: Amsterdam NL

Abstract

waLBerla is a modern open-source massively parallel multiphysics simulation framework with a focus on CFD applications. It uses the lattice Boltzmann method (LBM), which is an alternative to classical Navier-Stokes solvers for computational fluid dynamics simulations. waLBerla scales on some of the top clusters in the world due to carefully designed distributed data structures. As a showcase, we use particle-resolved sediment transport simulations using LBM and the discrete element method (DEM). In this scenario, the LBM simulation dominates the computational cost. In our implementation, the LBM simulation runs on the GPU and the DEM simulation on the CPU. We analyze the performance on a heterogeneous compute cluster.

Authors with CRIS profile

Samuel Kemmler Lehrstuhl für Informatik 10 (Systemsimulation) Harald Köstler Lehrstuhl für Informatik 10 (Systemsimulation)

How to cite

APA:

Kemmler, S., & Köstler, H. (2023). Walberla: a Multi-Physics Open-Source Software Framework for Scalable and Efficient CFD. Paper presentation at SIAM Conference on Computational Science and Engineering, Amsterdam, NL.

MLA:

Kemmler, Samuel, and Harald Köstler. "Walberla: a Multi-Physics Open-Source Software Framework for Scalable and Efficient CFD." Presented at SIAM Conference on Computational Science and Engineering, Amsterdam 2023.

BibTeX: Download